3394
Journal of the American Chemical Society
generation of thiocystine, CySSSCy, from the reaction of cystine with the unstable cysteine hydropersulfide has been noted.I4 The latter is obtained enzymatically by the action of cysteine desulfhydrase on cystine.l5 The reconstitution of parsley apoferredoxin in the presence of rhodanese (thiosulfate-cyanide sulfur transferase), dithiothreitol, and thiosulfateI6 suggests that the sulfur transferase activity of rhodanese also could be involved in ferredoxin biosynthesis. The equilibrium (eq 1) that describes the “sulfur-rich” rhodanese (E.&) action on dihydrolipoate has been reported. E 3 2 21ip(SH)2 e 21ipS2 2HS2H+ E (1) I n the course of formation of lipoate, lipS2, a chromophore was observed and assumed to be a n organic persulfide intermediate. Whether this intermediate, with a hydropersulfide functional group, Iip(SH)(SSH), undergoes further reaction with lipoate (eq 2) with formation of a cyclic lipoate trisulfide remains to be established. Such a reaction, however, would be analogous to the reaction reported to occur between CySSH and C Y S S C Y(eq ’ ~ 3). A suggestion is advanced, that linear or cyclic trisulfides, upon reduction, may serve as sources for inorganic sulfide. A synthesis of cyclic trithiothreitol was undertaken to examine the ability of this molecule to act as an oxidizing agent for the [Fe(SC6Hj)4] 2- complex. TrithiothreitolI8 reacts readily with [Fe(SC6H5)4I2- in D M F to form [ F ~ ~ S ~ ( S C ~ H which S ) ~ ] was ~ - , obtained in good yields and characterized by comparison with an “authentic”
+
+
+
+
+
+
lip(SH)(SSH) lips2 lips3 lip(SH)2 (2) CySSH CySSCy @ CySSSCy C y S H (3) Acknowledgment. The financial support of this project by a grant (CHE-75-23495) from the National Science Foundation is gratefully acknowledged.
+
-+
+
SupplementaryMaterial Available. Observed structure factors for [(C,jH5)4P]2Fe2S12(9 pages). Ordering information is given on any current masthead page. References and Notes (1) D.G. Hoiah and D. Coucouvanis, d Am. Chem. Soc.,97, 6917 (1975). (2) D. Coucouvanis. D. Swenson, N. C. Baenziger, D. G. Holah, A. Kostikas, A. Simopoulos, and V. Petrouleas, J. Am. Chem. Soc., 98, 5721 ( 1976). (3) (a) A. Kostikas, V. Petrouleas. A. Simopoulos. D. Coucouvanis, and D. G. Holah, Chem. Phys. Lett., 38, 582 (1976); (b) N. Sfarnas, A. Simopoulos, A. Kostikas, and D. Coucouvanis. unpublished work. (4) C. E. Johnson, J. Phys. (Paris), 35, CI-S7 (1974). (5) K. D. Watenpaugh, L. C. Sieker, J . R. Herriot, and L. H. Jensen, Acta Crystallogr., Sect. B, 29, 943 (1973). (6) R. H. Holm and J. A. lbers in “Iron Sulfur Proteins”, Vol. 3, W. Lovenberg. Ed.. Academic Press, New York, 1977, Chapter 7, and references therein. (7) J. J. Mayerle, S. E. Denmark, 8. V. De Pamphiiis, J. A. Ibers, and R. H. Holm, J. Am. Chem. SOC.,97, 1032 (1975). (8) The nature of this reaction is presently under study. (9) K. A. Hofmann and F. Hochtlen, Chem. Ber., 36,3090 (1903). (10) H. Kopf, B. Block, and M. Schmidt, Chem. Ber., 101, 272 (1966). (11) A. E. Wickenden and R. A. Krause, lnorg. Chem., 8,779 (1969). (12) P. E. Jones and L. Katz, Acta Crystallogr., Sect. 8, 25, 745 (1969). (13) The strictly inorganic Fe2Sj22- anion undergoes interesting sulfur abstraction reactions with triphenylphosphine which are presently under study. (14) T. W. Szczepkowski and J . L. Wood, Biochim. Biophys. Acta, 139, 469 (1967). (15) D. Cavallini, B. Mondovi, C. De Marco, and A. Scioscia-Santoro, Enzymologia, 24, 253 (1962). (16) U. Tomati, R. Matarese, and G. Federici, Phytochemistry, 13, 1703 (1974). (17) M. Villarejo and J. Westley, J. Biol. Chem., 238, 4016 (1963). (18) A white crystalline solid of the composition C4H802S3was obtained by the reaction of dithiothreitol with SC12 in ether. Analytical and NMR spectra data are consistent with the formulation of this molecule as a trisulfide.
D. Coucouvanis,* D. Swenson P. Stremple, N. C. Baenziger Department of Chemistry, University of Iowa Iowa City, Iowa 52242 Received January 22, 1979 0002-7863/79/lSOl-3392$01 .OO/O
/
101.1 2
/ June 6 , 1979
Peptide Synthesis in Water and the Use of Immobilized Carboxypeptidase Y for Deprotection Sir: We report here a method for peptide synthesis in aqueous systems in which coupling and deblocking are carried out a t p H 6 and 8.5, respectively (Scheme I). Water-soluble carbodiimide and amino acid ethyl esters are used to elongate the chain from the carboxyl end (Scheme 11). The removal of the ethyl ester blocking group is accomplished by treatment with immobilized carboxypeptidase Y (CPY) a t pH 8.5 and room temperature. The bound enzyme is removed by filtration and the filtrate used directly for the addition of the next amino acid. Peptides were grown on carboxymethyl-poly(ethy1ene glycol)-glycylmethionine (CM-PEG-Gly-Met). The use of this handle provides solubility for the growing chain and permits facile release of finished peptide using CNBr.’ The advantages of a polymeric handle have been pointed out by Mutter et al., who used underivatized PEG in synthesis of peptides by extension of the amino terminus in organic solvents.2 The preparation of chemically and optically pure H-Leu-Phe-Leu-OEt is illustrated below. C P Y is an exopeptidase from yeast with a very broad speci f i ~ i t yThe . ~ enzyme also has esterase activity. An important point for this study is that the p H optima for the peptidase and esterase activities are quite different. In preliminary studies, we demonstrated the absence of peptidase activity in the time required to deblock Z-Leu-Phe-OEt. Immobilization of the enzyme on CL-Sepharose does not appear to perturb the p H profiles of the enzyme.s It is also significant that a D-amino acid a t the C terminus or the penultimate position prevents ester hydrolysis by CPY. This feature assures optical purity of the final product. PEG (14 g, mol wt 6000-7000, MCB) and potassium tertbutoxide (10 g, Aldrich) were dissolved in tert-butyl alcohol ( 1 50 mL) by warming to 40 OC. Ethyl bromoacetate (5 mL) was added over a period of 10 min. After we stirred the solution for 2 h, the solvent was evaporated. The residue was dissolved in 100 m L of 1 N N a O H . After 2 h a t room temperature, the pH of the mixture was adjusted to 2. The CM-PEG was extracted into CHC13 (two 200-mL portions). The organic extract was washed with water and dried with anhydrous Scheme I. Preparation of the Handle HOCH2CH20(CH2CH2O),,CH2CH20H
I
1)
(PEG)
BrCH2COOEt/KBuOt
2)
purification
3)
H y d r o l y s i s , pH 10.5
PEG-CH~COOH (CM-PEG)
1
J
H-Gly-OMe,
pH 6.0
l-Ethyl-3(3-dimethylaminopropyl I c a r b o d i i m i d e (EDC) 1) p u r i f i c a t i o n 2)
pH 10.5
CM-PEG-Gly-OH 1) H-Met-OEt, EDC 2)
pH 6 . 0
purification
I
CM-PEG-Gly-Met-OEt pH 8.5
1
Immobilized Carpoxypeptidase Y
(I-CPY)
CM-PEG-Gly-Met-OH
0 1979 American Chemical Society
3395
Communications to the Editor
nmol of Phe; sample 3 (hydrolyzed with immobilized aminoNapS04. Evaporation of the solvent yielded 12 g of CMpeptidase M8), 27.0 nmol of Leu and 13.5 nmol of Phe. PEG. We demonstrated the absence of peptidase activity during Gly-OEt-HCI (1.4 g) and CM-PEG (3.4 g) were dissolved the time required for complete deblocking of the ester, Zin 35 mL of water. The pH was adjusted to 6 with triethylamine. l-Ethyl-3-(3-dimethylaminopropyl)carbodiimide~HClLeu-Phe-OEt. As an additional test, we checked for peptidase activity on a very good CPY substrate attached to PEG. After (2.0 g) was added, and the pH was maintained by automatic complete deblocking of CM-PEG-Phe-Met-Tyr-OEt (-20 addition of 0.1 N HCI. After 3 h at room temperature, the min, pH 8.5,25 'C), a sample corresponding to 1000 nmol of reaction mixture was acidified to pH 2 and the product was tyrosine was applied to the amino acid analyzer. An insignifextracted into 250 mL of CHC13. The organic layer was icant amount of free tyrosine was detected (about 0.1%). washed with 1 N HCI and water. After removal of the CHC13, The question of racemization arises when extension of the the CM-PEG-Gly-OEt was saponified at pH 10.5 using a pH C terminus of the peptide chain is chosen as the approach to stat and 0.1 N NaOH. After acidification and extraction, the sequential synthesis. The generally accepted mechanism for product (3.3 g) had 100 nmol of glycine/mg according to racemization involves the deprotonation of an oxazolinone amino acid analysis. i ~ ~ t e r m e d i a tWe e . ~ reasoned that this deprotonation would not Met-OEt-HCI was reacted in a manner analogous to Glybe significant at pH 6 in water. Z-Leu-Phe-Leu-OEt made by OEt. CM-PEG-Gly-Met-OEt was deblocked by treatment our method had an optical rotation identical with that of a with immobilized CPY at pH 8.5. About 30 ATEE units of sample made by the excess mixed anhydride method. Our enzyme was used.3 The enzyme was immobilized on CLpeptides are chemically homogeneous at every stage, indicating Sepharose as previously d e s ~ r i b e d or , ~ by adsorption on imthat complete deprotection occurred. Base uptake during the mobilized concanavalin A with glutaraldehyde cross-linking.6 deprotection step also matched the predicted amount. Also, Deblocking of CM-PEG-Gly-Met-OEt took 5 h. However, agreement of amino acid ratios between analyses on samples as the chain length is increased, the deblocking time was draof H-Leu-Phe-Leu-OEt prepared by acid and enzymatic hymatically reduced, probably as a result of increased subsite drolysis strongly suggests that our method yields optically pure binding. CM-PEG-Gly-Met-Leu-Phe-Leu-OEt was prepared peptides. by successive addition of Leu-OEt, Phe-OEt, and Leu-OEt as For a more sensitive racemization test we adapted the described above. Starting with 1.75 g of CM-PEG-Glymethod of Kitada and Fujino,Io which is based on the sepaMet-OEt produced 1.50 g of CM-PEG-Gly-Met-Leu-Pheration of dipeptide diastereomers." With a 60 X 0.9 cm colLeu-OEt. Amino acid analyses, performed at each step using umn of Beckman AA-15 resin, L-Leu-L-Leu (60 min) was well a hydrolysate of the CM-PEG-peptide, indicated that couseparated from D-Leu-L-Leu (75 min). The conditions were pling and deblocking went to completion. Hydrolysis was as follows: flow rate, 1 mL/min; column temperature, 5 5 'C; performed according to Westall and Hesser.' buffer, 0.35 M citrate (pH 5.25). The method is capable of Release of the peptide was accomplished by treatment of detecting 1 nmol of D-L isomer in 200 nmol of peptide. CM-PEG-Gly-Met-peptide with CNBr at room temperature. CM-PEG-Phe-Leu-Leu-OH was prepared as described CM-PEG-Gly-Met-peptide (180 mg) was dissolved in 1 mL above. The dipeptide was released using a-chymotrypsin. Also, of distilled water. To this 0.018 mL of a CNBr stock solution two separate batches of PEG-Phe-Met-Leu-Leu-OH were (0.1 M in acetonitrile) was added. The mixture was lyophilized prepared and cleaved with CNBr according to ref 9. The after 1 h. Water ( 1 mL) was added, the pH of the mixture was maximum extent of racemization was 0.7% out of nine analyses adjusted to 2, and lyophilization was repeated. The residue was involving three distinct batches of peptide. dissolved in N a H C 0 3 buffer (pH 8 . 5 ) , and the solution was extracted with EtOAc. CNBr treatment of 180 mg of C M Another question is whether or not deblocking with CPY is applicable in the synthesis of peptides that involve side-chain PEG-Gly-Met-Leu-Phe-Leu-OEt yielded 4.5 mg of peptide protection. We have synthesized H-Ala-Ala-Cys(ACM)(64% based on glycine content of the handle). The product, Lys(Z)-OH by the method detailed above. Deprotection and H-Leu-Phe-Leu-OEt, was identical with a sample of authentic coupling went to completion at each stage as judged by TLC peptide as judged by TLC: Rf 0.79 (n-BuOH-AcOH-H*O and amino acid analyses. About 320 mg of PEG-Gly-Met4:1:5) and R, 0.85 (n-BuOH-AcOH-HlO-pyridine 15:3: Ala-Ala-Cys(ACM)-Lys(Z)-OH was obtained starting with IO: 12). The optical rotation of our product was similar to that 400 mg of PEG-Gly-Met-OEt. The peptide was released as of the standard: [ a ] 2 -0.45' 5~ (standard, 1% in EtOH) and described above (42% yield after extraction). The analytical -0.42 (product, 1% in EtOH). The amino acid analyses of the released peptide showed the following: sample 1 , 20.9 nmol of data of the released peptide follow (amino acid analysis, Leu and 10.4 nmol of Phe; sample 2 , 4 5 3 nmol of Leu and 21.9 nanomoles): Ala, 88.4; Cys(ACM), 42.2; Lysl, 44.4:TLC, R, 0.52 (H-BuOH-ACOH-H~O,4:1:5; silica gel).
Acknowledgments. We thank Pierce Chemical Co. and Eli Lilly Co. for financial support. We are indebted to Dr. M. A. Tilak for preparing Z-Leu-Phe-Leu-OEt via the excessmixed-anhydride approach.I2 G. P. Royer is a recipient of a Public Health Service Research Career Development Award (5-K-4GM-0005 1).
Scheme 11. Chain Elongation and Release CM-PEG-Gly-Met-OH H-AA1-OR 1) EOC, pH 6 2) purification
References and Notes CM-PEG-Gly-Met-AA1-OR pH 8.5 I-CPY
1
CM-PEG-Gly-Met-AA1-OH l CM-PEG-Gly-Met-[AA1. I
CNBr
i
pH
peptide
'
..
.AAnl-OH
(1) E. Gross, Methods Enzymol.. 25, 419-423 (1976). (2) M. Mutter, H. Hagenmaier, and E. Bayer, Angew. Chem., 1nt. Ed.,Engl., 10, 811-812 (1971). (3) R . Hayashi, Methods Enzymol., 45, 569-587 (1976). (4) Y. Nakagawa and E. T. Kaiser, Biochem. Biophys. Res. Commun., 61, 730-734 (1974). (5) F. A. Liberatore, J. E. Mclsaac, Jr.. and G. P. Royer. FEBS Lett., 68, 45-48 (1976). (6) H. Hsiao and G. P. Royer, manuscript in preparation. (7) F. Westall and H. Hesser, Anal. Biochem., 61, 610-613 (1974). (8) G. P. Royer and J. P. Andrews, J. 6/01. Chem., 248, 1807-1812 (1973). (9) M. Bodanszky, Y. S. Klausner, and M. A. Ondetti, "Peptide Synthesis", 2nd ed., Wiley, New York. 1926, p 137.
3396
/ 101:12 / June 6, 1979
Journal of the American Chemical Society
1A
(10) C. Kitada and M. Fujino, Chem. fharm. Bull., 26,585-590 (1978). (11) J. M. Manning and S . Moore, J. Bo/. Chem., 243, 5591-5596 (1968). (12) M. A. Tilak, Tetrahedron Lett., 849 (1970).
I
0.2
Garfield P. Royer,* G . M. Anantharmaiah Department of Biochemistry, The Ohio State University Columbus, Ohio 43210 Received June 12, 1978
Kinetics of CO Binding to Fe(TPP)(Im) and Fe(TPP)(Im-): Evidence Regarding Protein Control of Heme Reactivity
021
B
Sir:
Protein structure modulation of the reactivity of a heme prosthetic group has been a focus for numerous investigations.’ For example, ligand binding kinetics of hemoglobin are affected by protein conformation.1a,2The CO on-rate for the low-affinity T state is -20-60-fold less than that for the high-affinity R state.3 One plausible mechanism of this modulation has been advanced by Peisach and collaborator^.^ It is generally proposed that a protein can induce changes in heme reactivity through structural changes a t the proximal imidazole. For hemoglobin, in the low-affinity state the proximal histidine is thought to be in its neutral form with a proton on N - 1. The high-affinity form is considered to have a strong hydrogen bond to that proton, which in the limit might be thought of as corresponding to a deprotonated imidazole as the sixth ligand. Such a mechanism also has been invoked to discuss the electronic structure of cytochromes c from different organism^,^ and the crystal structures of a number of hemoproteim6 Prompted by our previous work on the mechanisms of heme-protein interaction^,^,' we now have tested the relation of heme reactivity and the imidazole protonation state by the direct comparison of the C O binding rate, for a five-coordinate ferrous porphyrin model with neutral imidazole as fifth ligand, Fe(P)(lm),x with that for the model with deprotonated imidazole as the fifth ligand, Fe(P)(Im-). Deprotonation can indeed alter the CO binding rates to a degree in excess of the difference in binding rates to the T and R states of Hb. However, the difference is in the opposite sense to that which would normally be expected: deprotonation decreases the rate of C O binding. Kinetic measurements were made by monitoring absorbance changes after flash photolysis of an Fe(TPP)(B)(CO), B = Im, Im-, s ~ l u t i o n , ~ using - ~ * a computer-interfaced apparatus of conventional design.13 The Fe(TPP)(Im)(CO) complex is highly photolabile and the photoproduct rebinds CO with a rate constant, kobrd a [ c o ] . However, the difference spectrum of Fe(TPP)(Im)(CO) and the product obtained immediately after the flash (-20 ps) does not correspond to the formation of Fe(TPP)(lm), which should have a Soret maximum at -435 nm.” Rather, the kinetic difference spectrum is the same as the static difference spectrum [ Fe(TPP)( Im)2] - [Fe(TPP)(Im)(CO)], obtained by direct subtraction of the absorbance spectra of the appropriate complexes, which is also presented in Figure IA. This indicates that a second Im is bound within the flash lifetime. Therefore the overall stoichiometry of the CO rebinding reaction as observed in the regeneration of Fe(TPP)(Im)(CO) after the photolysis flash is ligand replacement Fe(TPP)(Im)2
+ CO
-
Fe(TPP)(Im)(CO)
+ Im
and not the ligand addition reaction Fe(TPP)(Im)
+ CO 2Fe(TPP)(Im)(CO) 0002-7863/19/1501-3396$01 .OO/O
420
440
460
X (nm)
Figure 1. (A) Static difference spectrum, Fe(TPP)(lm)2-Fe(TPP)(Im)(CO), -; kinetic difference spectrum after flash photolysis of Fe(TPP)(lm)(CO), -. (B) Static differencespectrum, Fe(TPP)(lm-)*Fe(TPP) (Im-) (CO),i -; kinetic difference spectrum after flash photolysis of Fe(TPP)(lm-)(CO), .-.
Scheme I
K,
k,
I
k,
+CO k’
Fe(C0)
I
+CO
Fe (B)(CO)
+B which would correspond to the CO binding reaction by Hb. This observation is expected from previous studies, such as those of Traylor;I4 binding of CO to an iron(I1) porphyrin in the presence of excess base (B) can usually be described in terms of a preequilibrium between the Fe(B), ( n = 0, 1, 2),11314 where we have suppressed the porphyrin abbreviation (Scheme 1). Here Kl and K2 are the measured static binding constants, and k4 and k5 are the second-order CO binding rates, of which k5 is the quantity of interest; the reverse reaction, loss of CO, can be neglected on our time scale. When comparing rates for different bases, we will keep track by writing k5(B). Under the conditions of our experiments, the results of White et al.14cshow that C O addition is rate limiting, and that the ligand rebinding rate obeys an equation kobsd/[CO] = k 4 / z -k k ~ ( K i [ B l / z )
+
+
(1)
where Z = 1 KIIB] KIK2[BI2.The equilibrium constants for binding Im by Fe(TPP) are known: Kl = 8.8 X I O 3 M-l and K2 = 7.9 X 1 0 4 M-I.” We performed a series of rate measurements a t constant [CO] and with varying [Im], and, from the plot of kobsd2/[CO] = k 4 ksKl[B] vs. [B] (Figure 2 ) , we obtain k4 from the interecept and k5 from the slope (Table I). The Fe(TPP)(Im-)(CO) complex is also photolabile, and
+
0 1979 American Chemical Society
